[maker-devel] Questions about re-annotating
Carson Holt
carsonhh at gmail.com
Tue Jun 23 10:46:58 MDT 2015
Hi Malcolm,
If you are not running with any new evidence (EST or protein), and you are passing in all previous evidence as GFF3, then you don’t need to run the repeat masking steps if you don’t want (because you evidence is now static and won’t be affected). You could also just rerun in the same directory as a previous MAKER run as well (MAKER will reuse archived raw analysis files where appropriate). I prefer the second option to GFF3 pass-through when available.
The --again option is really only to force MAKER to rerun something if there is an error. It won’t rerun the individual analysis steps, rather it will just re-parse saved raw reports. This may be done because of IO issues on clusters that can be common on network mounted file systems (i.e. when you try and read and write from 2,000 CPUs simultaneously weird things sometimes happen).
Exonerate will always run if you provide ESTs or proteins as FASTA. It will never run if you provide them as GFF3. The est2genome option just tells MAKER to try and turn polished EST alignments directly into genes (without modification). It does not cause exonerate to run. This is generally just used to get a set of preliminary models that are sufficient for training gene predictors.
Thanks,
Carson
> On Jun 23, 2015, at 12:46 AM, Cook, Malcolm <MEC at stowers.org> wrote:
>
> Hi,
>
> I am re-running an analysis after having adding `est_gff=/path/to/cuffmerge.gff3` and want to ensure the least amount of work is re-done.
>
> When I re-run maker, should I provide any additional options. In particular, either "--again" ?
>
> On this subject, I read in http://weatherby.genetics.utah.edu/MAKER/wiki/index.php/Updating_annotations_in_light_of_new_data
>
> "if you are using est_gff= you don't even have to repeat mask." - does this mean I should always run maker with option -RM_off when updating annotation?
>
> On a related note, if I have est2genome turned on, what changes to configuration should cause exonerate to run? I'm guessing:
> adding new est= files
> turning on alt_splice=1
>
> Thanks!
>
> ~ malcolm_cook at stowers.org
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
More information about the maker-devel
mailing list